We have previously shown that the coexpression of c-myc and transforming growth factor alpha (TGF-alpha) as transgenes in mouse liver results in major enhancement of neoplastic development in this organ as compared with expression of either of these transgenes alone. Despite morphological similarities in the sequence of events between the two transgenic lines, the dramatic acceleration, extent, and severity of hepatic lesions in c-myc/TGF-alpha mice clearly demonstrated the synergistic effects of this transgenic combination. Although c-myc/TGF-alpha and c-myc females displayed longer latency and lower tumor incidence, the pathological changes were the same as those seen in the male mice, including the formation of hepatocellular carcinomas (HCC), which are absent in TGF-alpha single transgenic females. Tumors in single- and double-transgenic mice showed induction of the endogenous c-myc and TGF-alpha and, most frequently, unchanged or decreased epidermal growth factor receptor, further indicating the collaborative role of c-myc and TGF-alpha in providing a selective growth advantage to tumor cells independently of the epidermal growth factor receptor levels. To identify possible tumor precursors, we focused particularly on the dysplastic changes preceding and accompanying the appearance of preneoplastic and neoplastic lesions in the double-transgenic mice. Early on, these changes were characterized by the appearance of large dysplastic hepatocytes, mostly pericentrally, expressing high levels of TGF-alpha and uPA, as well as TGF-alpha1, particularly in apoptotic cells. Transplantation of the transgenic liver tissues harboring only dysplasia with or without vascular lesions onto nude mice was able to yield HCCs composed of small diploid cells, suggesting that initiated cells are generated during the early dysplastic phase and can progress to HCC. It is therefore likely that large dysplastic hepatocytes undergo apoptosis, which may be closely associated with the "up-regulation" of TGF-131 and uPA, whereas other cells evolve into the precursor population for HCC. Due to the simultaneous presence of c-myc, TGF-alpha, and dysplasia in premalignant human liver diseases, our transgenic mouse system appears to be an appropriate model for studying human hepatocarcinogenesis.